1. Field of the Invention
Methods consistent with the present invention relate to refining of node position information in a wireless sensor network. More specifically, in a wireless sensor network having an ultra wideband communication network and a wireless personal area network where the initial positions of respective nodes are set, the present invention relates to a method of refining node position information to refine the position of nodes based on a distance measured between the respective nodes, utilizing nodes at a 2-hop distance, or more than a 2-hop distance, when the number of 1-hop neighboring nodes is not sufficient to perform triangulation.
2. Description of the Related Art
There are generally three ways of position estimating in a multi-hop wireless sensor network or a wireless personal area network.
Among the three general ways of position estimating, the first one estimates distance between an anchor node and an unknown node at a multi-hop distance according to the measured inter-hop distance, and then estimates the position of the unknown node using range-based position estimating such as sum-dist. The first method provides advantages such as low estimation error in an environment where there is a low possibility of a distance measuring error. However, this method has a drawback that it is heavily dependent on the accuracy of distance measurement.
The second method for position estimating is a range-free estimation, which estimates distance between an anchor node and an unknown node using computed distance instead of measured distance, and then estimates the position of the unknown node. This second method provides the advantage that the accuracy of position estimating is hardly influenced by distance measurement errors. However, it is hard to find an accurate position of the node.
The third way of position estimating is to use advantageous features of the above two methods, and this is a multi-modal method such as Parametric Learning-based Distance Location (Para-LDL). The Para-LDL method can provide far better performance than the other estimating methods, but still has the problem of a distance measurement error which is generated as the unknown node is at a multi-hop distance from the anchor node in a multi-hop sensor network.